Tick Fever in the Northern Beef Industry:

Transcription

1 Tick Fever in the Northern Beef Industry: Prevalence, Cost/ Benefit of Vaccination, Considerations for Genotypes, Livestock Management and Live Cattle Export Project number DAQ.107 Final Report prepared for MLA by: R E Bock Department of Primary Industries Queensland Meat and Livestock Australia Ltd Locked Bag 991 North Sydney NSW 2059 ISBN May 1999 MLA makes no representation as to the accuracy of any information or advice contained in this document and excludes all liability, whether in contract, tort (including negligence or breach of statutory duty) or otherwise as a result of reliance by any person on such information or advice. MLA 2004 ABN Info Category

2 EXECUTIVE SUMMARY Cattle producers in Northern Australia need more information on the economic significance of tick fever in extensive cattle herds and the benefit to cost ratio of vaccination programs. While well documented for Bos taurus cattle, the factors that influence the eventual outcome of tick fever infection are not well understood for Bos indicus cattle and their crosses. Meat and Livestock Australia was approached by key northern producers to help address this deficiency and DAQ107 was developed and carried out by QDPI. This project aims to present the information that is available and provide new information to assess the risk of losses due to tick fever in northern herds and to define the benefit to cost of tick fever vaccination. 1

3 Table of Contents 1. METHODOLOGY KEY FINDINGS RECOMMENDATIONS FOR CONTROL OF TICK FEVER IN NORTHERN AUSTRALIA Tick fever vaccination Quality assurance system APPLICATION Vaccination Decision support service Sentinel herds GENERAL DETAILS Project participants Project leader Project investigators Modelling... Error! Bookmark not defined Collaborators OBJECTIVES Expected outcomes BACKGROUND Field evidence of tick fever in NW Queensland Tick fever vaccine Factors involved in the occurrence of tick fever

4 7.3.1 Transmission of tick fever Breed resistance to tick fever Endemic stability Age resistance Models of endemic stability and disease prediction Mahoney model Ramsay s disease prediction/vaccination and discounted cashflow analysis model Disease prediction model Assumptions of the disease prediction model Vaccination strategies Vaccine coverage and efficacy Discounted cash flow analysis Project objectives and findings CONCLUSIONS Key findings Recommendations for control of tick fever in Northern Australia In endemic areas In live cattle export trade Application Vaccination Decision support service Sentinel herds...33 ACKNOWLEDGMENTS...33 REFERENCES

5 APPENDIX 1: WORK PROGRAM/MILESTONES (NON- BUDGET)...39 APPENDIX 2: INPUT PARAMETERS FOR RAMSAY MODEL...40 APPENDIX 5: SPREADSHEETS IN THE MODEL...43 APPENDIX 6: COMMUNICATION

6 1. METHODOLOGY The project established the seroprevalence of the three tick fever parasites (A marginale, B bovis and B bigemina) in cattle by cross-sectional surveys on cattle properties in north-west Queensland over a period of 3 years. B indicus and crossbred cattle from north-west Queensland were moved to a highly endemic area and monitored to estimate losses due to tick fever in a worst case scenario. The information from the surveys and natural exposure as well as expert opinion was used in a spreadsheet model to produce disease prediction and production loss information linked to discounted cashflow analysis. 2. KEY FINDINGS Endemic stability was not observed in any of the shires surveyed for any of the 3 tick fever parasites and large numbers of cattle in the region are naive for the three tick fever parasites. The exposure trials showed that B indicus cattle are very resistant to both babesia parasites, but they and crossbred cattle are very susceptible to A marginale. Crossbred cattle are resistant to B bigemina but B bovis will cause significant mortalities in crossbreds. Increasing the B taurus content of cattle will substantially increase the risk of tick fever losses. The spreadsheet model to predict the benefit cost of vaccination has been tested using this information. The producer s attitude to risk, as well as long-term view of markets and breed structure have to be considered, but the model allows producers to assess the risk more accurately for their particular herd and therefore make more informed decisions on tick fever control. 3. RECOMMENDATIONS FOR CONTROL OF TICK FEVER IN NORTHERN AUSTRALIA 3.1 Tick fever vaccination In balance, it is our opinion that northern cattle producers should carry out an annual weaner tick fever vaccination program. This is because of: low natural transmission and therefore poor naturally acquired immunity; the need to protect overseas live cattle market access which can be denied for 6 to 12 months after a tick fever outbreak on individual properties; high susceptibility of B indicus and cross-bred cattle to anaplasmosis; the trend to increase the B taurus herd content. The bivalent vaccine (B bovis and A centrale) can be considered for B indicus cattle, but to ensure broad-spectrum protection, trivalent vaccine (B bovis, B bigemina and A centrale) should be used especially in cross-breds and B taurus cattle. 5

7 3.2 Quality assurance system Currently, when live cattle export protocols require tick fever vaccination it has to be given near the time of shipment. Because of the age of and added stress to these cattle they are more likely to have severe vaccine reactions during shipping, quarantine or feedlotting and this may impact on their performance. These protocols should be amended to accept vaccination at weaning. This age group has far fewer vaccine reactions and there is the added benefits of protection from weaning to shipment and improved cattle quality at the destination. An on-farm quality assurance system that will ensure proper use of the vaccine and identification of immunised cattle that is also accepted by the importing country would be required, but the benefits would be considerable. 4. APPLICATION 4.1 Vaccination The Tick Fever Research Centre (TFRC) produces tick fever vaccine that is readily available for use in extensive areas. Producers must assess the risk of tick fever from the information available and make a decision on vaccination that is consistent with their own attitude to risk. If supply of chilled vaccine is a problem in the more remote areas they can use the frozen tick fever vaccine which has a five-year shelf life stored in liquid nitrogen. 4.2 Decision support service Using the information generated from this project and the spreadsheet model it is possible to offer a decision support service tailored to individual herds. We suggest that this be operated on a fee for service basis which would entail the serological testing of approximately 50 weaner age cattle from a herd or in the case of larger properties a number of subgroups to adequately assess the current seroprevalence on the property. This information along with herd data would then be entered into the model to estimate the return on vaccination costs. Various scenarios looking at changes in the risk over the 8-year outlook of the model can be demonstrated to allow a more complete picture and give some information on the sensitivity of the output to scenario changes. 4.3 Sentinel herds Annual serological tests on weaners from sentinel herds in specific shires of interest would give an estimate of changes in risk over time. This information processed through the spreadsheet model could be used to advise producers in a shire on possible changes in tick fever risk. 5. GENERAL DETAILS Start Date: 01 July 1995 (Actual January 1996) End Dates - Proposed: 31 December 1998 Actual: 31 March

8 5.1 Project participants Project leader Mr R.E. Bock, Principal Scientist Tick Fever Research Centre 280 Grindle Road Wacol, Queensland 4076 Telephone: , Fax , Project investigators TFRC Mr AJ De Vos, Centre Coordinator. Mr TG Kingston, Research Station Manager. Mrs S Singh, Laboratory Technician. West Region Mr WP Lehmann, Senior Stock Inspector, QDPI, Julia Creek (now Blackall) Mr AC Rayner, Rural Services Coordinator, QDPI, Longreach QDPI Stock Inspectors involved in sampling and extension programs S Anderson S Laffey R Pearce E Beekhuizen A Lipp I Purvis T George D Marshall B Torenbeek D Healy B Mason T Vinson P Jordan G McDougall D Hogarth Modeling Dr G Ramsay (Formerly QDPI) Veterinary Epidemiologist Secretariat for the Pacific Community Private Mail Bag Suva, Fiji Collaborators Mr N Haug Manager Research and Development Australian Agricultural Company GPO 587 Brisbane, QLD 4001 Mr J Armstrong Stanbroke Pastoral Company GPO Box 155 Level Creek St Brisbane QLD

9 6. OBJECTIVES Perform survey of the current prevalence of the three tick fever parasites (A marginale, B bovis and B bigemina) in cattle on a cross section of properties in northern cattle industry over a period of 3 years. Estimate the losses due to tick fever in the breeds currently utilised on selected properties in the northern cattle industry. Provide the northern producers with information that will allow them to make rational decisions regarding breed type, vaccination strategy, tick control and cattle movements in relation to the management of tick fever. 6.1 Expected outcomes The project will yield information on tick fever in North Queensland as follows: Distribution Spread Immune status of herds (Export quality assurance) Genotype Susceptibility Cost/Benefit analysis of control This should lead to: Better understanding of conditions under which tick fever can be maintained in N Qld (effect of genotype, location, droughts) and this information can be extrapolated to others areas of Northern Australia. Estimation of potential losses if tick fever occurs in Northern Australia Assessment of risk in existing breeds in Northern Australia 7. BACKGROUND The cattle tick, Boophilus microplus, and tick fever parasites, Babesia bovis, Babesia bigemina and Anaplasma marginale are endemic in Northern Australia. However, cattle tick numbers have evidently been declining in north-west Queensland in recent years. Cattle ticks were not seen on many of the properties within the officially tick infested parts of northwest Queensland from 1990 to 1996 (AC Rayner, unpublished observations). Limited serological surveys in north-west Queensland have failed to find evidence of Babesia transmission on several properties previously known to have cattle tick (Queensland Department of Primary Industries structured surveillance 1990 to 1995, unpublished). This reduction in tick numbers is presumed to be due to the combined effects of resistant cattle and recent droughts. Two surveys (O Rourke et al 1992; Bock et al 1995) showed that virtually all beef producers in far Northern Australia are using cattle with greater than 3/8 Bos indicus infusion. 8

10 Mahoney et al (1981) in a study comparing B bovis transmission in Bos taurus and cattle with 3/8 to ½ B indicus infusion in south-east Queensland concluded that, in an environment unfavourable for tick survival, stocking with crossbred cattle would lead to the virtual disappearance of the ticks. However, if the area was then occupied with pure B taurus cattle their low tick resistance more than compensated for environmental effects on the tick reproduction and the tick population flourished. Mahoney et al (1981) further predicted that tick resistance in crossbred cattle combined with the finding that they produce fewer B bovis infected tick larvae under similar levels of tick infestation to B taurus cattle might lead to the eventual disappearance of this parasite from crossbred herds. Few producers on extensive cattle producing enterprises in Northern Australia vaccinate their herds against tick fever (O Rourke et al 1992; Bock et al 1995). These producers are posing a number of questions that need to be addressed: (a) what will be the impact of tick fever on different breeds of cattle if ticks move back into these areas? (b) are the cross-bred cattle being exported to South-east Asia sufficiently resistant or immune to withstand the possible tick fever challenge levels at their destination (c) what is the cost/benefit of control options in different cattle breeds in the region? Simply put, they are asking: is tick fever of economic significance in extensive cattle herds in north-west Queensland, Northern Territory and Kimberley s region and should producers be vaccinating? The factors that can influence the eventual outcome for tick fever infection is fairly well documented for Bos taurus cattle. However, there are deficiencies in our knowledge of tick fever in Bos indicus cattle and their crosses. Meat and Livestock Australia was approached by key northern producers to help address the issues and DAQ107 was developed. The aim of this project is to present the information that is available and provide new information to assess the likely risk of losses due to tick fever in northern herds and the question of the need for vaccination. 7.1 Field evidence of tick fever in NW Queensland The QDPI maintains records of all laboratory submissions and diagnoses with an on-line information system (LOIS). These laboratory records show only 11 outbreaks of tick fever were confirmed in north-west Queensland from January 1990 to December The histories are given in Table 1 and they indicate that the parasites are present, but there are very few submissions. Table 1: Laboratory records from January 1990 to December 1998 showing only 11 outbreaks of tick fever confirmed in north-west Queensland by QDPI laboratories. No. Shire Aetiology At risk No sick No dead Breed Sex 1 Burke A marginale 2700? 200 B indicus female 2 Burke B bigemina? 1 0 B indicus female 3 Carpenteria B bovis 960? 30 B indicus female 4 Carpenteria B bovis Charolais male 5 Carpenteria B bigemina? 1 0 Santa male 6 Carpenteria B bovis B indicus female cross 7 McKinley B bigemina Santa and female Braford 8 Dalrymple B bovis??? B indicus female 9 Dalrymple A marginale??? B indicus female 10 Dalrymple B bovis 800 1? Shorthorn male 11 Dalrymple B bovis X-bred steers 9

11 Outbreak 3 involved cattle being mustered and held in a bullock paddock as part of a tuberculosis destocking program. Cases occurred in the bullock paddock, on route to a feedlot in the Beaudesert area and in the feedlot. Outbreak 4 is interesting because it involved naive Charolais bulls introduced to a property in Carpenteria shire that ran high content B indicus cattle. The manager had not noticed ticks on the property for several years, but with the arrival of the Bos taurus bulls ticks were soon noticed and so was tick fever. Outbreak 6 was in crossbred heifers and both B bovis and A marginale were confirmed in some cases with B bovis predominating. Later on Salmonellosis was confirmed in cohorts being fed in yards so the mortalities may not all be due to tick fever. Outbreak 7 was in cattle vaccinated with bivalent vaccine that were droving the stock routes having originated from a cattle tick free area. As most of these outbreaks were only confirmed because a stock inspector or company veterinarian was on hand, it is also probable that on these extensive properties many more limited outbreaks are occurring which remain undetected. A similar situation is apparent in the tick infested areas of Western Australia and the Northern Territory (Table 2 and 3). Table 2: Laboratory records from January 1990 to December 1998 showing only 10 outbreaks of tick fever confirmed in the Kimberleys region of Western Australia by Agriculture WA laboratories (From Dr Jeremy Allan personal communication) No. Region Aetiology At risk No sick No dead Breed Sex 1 West B bigemina Shorthorn steers Kimberley 2 East Babesia spp 158? 14 B indicus steers Kimberley 3 Kununurra B bovis Friesian female 4 Kununurra B bovis ? steers 5 Kununurra B bigemina B indicus steers 6 Kununurra B bigemina B indicus female 7 Kununurra B bigemina B indicus female 8 West B bovis B indicus steers Kimberley 9 West B bovis Shorthorn steers Kimberley 11 West Kimberley B bovis Shorthorn female Table 3: Laboratory records from January 1995 to December 1998 showing 5 outbreaks of tick fever confirmed in the Northern Territory by DPIF laboratories (From Dr Diana Pinch personal communication) No. Region Aetiology At risk No sick No dead Breed Sex 1 Darwin A marginale 19? 1 B indicus steers cross 2 Darwin A marginale 20? 0 B indicus female cross 3 Darwin A marginale 150? 9 B indicus female 4 Katherine Babesia spp Shorthorn steers 5 Katherine Babesia spp Friesian female Table 4 provides summary information from LOIS on tick fever outbreaks in the period from January 1990 to July 1996 for the whole of the State. It must be remembered that this is a biased dataset, many outbreaks go unconfirmed, data is not always complete and specimens are usually taken at the height of the outbreak so mortality figures are incomplete. However, the information gives some indication of the relative importance of the three tick fever parasites and indicates that, while far less common, tick fever outbreaks do occur in Bos 10

12 indicus herds. Approximately 65% of cattle herds in the cattle tick infested area of Queensland are greater than 3/8 Bos indicus (O Rourke et al 1992) and 77% of cases were from Brisbane/Moreton and Wide Bay/Burnett Australian Bureau of Statistics regions. Table 4: All Tick fever cases confirmed at QDPI laboratories from 1/1/90 to 30/6/96 Breed A marginale B bigemina B bovis Grand Total Percentage Bos indicus % Bos indicus cross % Bos taurus % mixed % X-bred* % unknown % Grand Total Percentage 11.0% 7.5% 81.5% * It is not always known what submitters define as X-bred but a large percentage of these can be expected to have at least 3/8 Bos indicus infusion. In a 1992 survey (Bock et al 1995) of Queensland beef producers in the B microplus infected area producers were asked to rank animal health factors reducing production in their herds based on mortality, reduced growth rates, carcass damage, treatment of disease and preventive measures. In North Queensland Botulism, Buffalo fly infestation, Ephemeral fever and plant poisoning were considered the most important problems. Tick fever did not rate as a problem in North Queensland. 7.2 Tick fever vaccine The Tick Fever Research Centre (TFRC) of the Queensland Department of Primary Industries (QDPI) currently produces the only tick fever vaccines in Australia. Standardised vaccines are produced using splenectomised donor calves to obtain large numbers of parasites with predictable, lower levels of virulence (Callow and Dalgliesh 1980; Callow 1984). The vaccines are based on live attenuated strains of B bovis and B bigemina and on A centrale, a usually benign parasite that induces partial immunity to A marginale. The A centrale organism was imported from South Africa in 1934 (Rogers and Shiels 1979). The vaccines are supplied as chilled trivalent (B bovis, B bigemina and A centrale) or bivalent (B bovis and A centrale) vaccines. The chilled vaccine has a shelf-life of only 4 days. The production of chilled monovalent (B bovis) vaccine was discontinued in The vaccines are sold on a cost recovery basis. TFRC also produces a frozen vaccine stored in liquid nitrogen and using glycerol as the cryoprotectant (Dalgliesh et al 1990). Vaccine is prepared by thawing cryovials in water at 37oC and then diluting the contents 1 in 5, in 1.5M glycerol PBS solution containing glucose and antibiotics. The contents of different vials can be mixed in the diluent to provide mono, bi or trivalent vaccine as desired. The vaccine has a shelf-life of 5 years in liquid nitrogen, but the thawed vaccine has to be used within 8 hours of thawing. Storage and handling requirements, the need to use the vaccine within 8 hours of thawing and the ready availability of easy-to-use chilled vaccine has restricted the frozen products acceptance in Australia. It is ideal for overseas markets and areas of Australia where the short shelf-life of chilled vaccine makes use of this product impracticable. On the whole, B bovis vaccine consisting of one attenuated strain affords sound protection against field challenge. Results of a field trial on five Queensland properties suggested that 11

13 vaccinated cattle were 15 to 16 times less likely to contract babesiosis than unvaccinated cattle (Emmerson et al 1976). Failures are uncommon and are likely to be due to an animalrelated ability to mount an immune response, immunogenicity of the strain of B bovis used in the vaccine or vaccine viability (Bock et al 1995). Time-dependency of immunity does not seem to be a factor and is discussed in section 5.3. Indications are that a single inoculation of B bigemina provides adequate to excellent protection against challenge. A centrale provides partial, variable protection against A marginale challenge. Protection against challenge in Australia is adequate and use of the vaccine in this country is justified. In a 1992 survey (Bock et al 1995) of Queensland beef producers in the B microplus infected area it was found that only 15% of North Queensland cattle producers used tick fever vaccine. 7.3 Factors involved in the occurrence of tick fever Transmission of tick fever Babesiosis Both B bovis and B bigemina are transmitted transovarially with a further period of development taking place in the larval stage and maturation in the salivary glands after tick attachment before transmission to the host. B bigemina B bigemina is transmitted by late nymphal and adult stages of B microplus (Hoyte 1961). Transmission can occur throughout the rest of the nymphal stage and by adult females and males (Callow and Hoyte 1961; Riek 1964; Dalgliesh et al 1978). Interhost transfer of these ticks can lead to a much-shortened prepatent period (6-12 days), but it is usually days after tick attachment (Callow 1984). B bovis is transmitted by the larval stage of Boophilus microplus only and infective forms are present in the salivary glands 2-3 days after attachment (Riek 1966). This period can be even shorter if larval ticks are exposed to temperatures above 30 o C on the pasture (Dalgliesh et al 1979). Prepatent periods are generally 6-12 days and peak parasitaemias are reached about 3-5 days after that (Callow 1984). This means that unvaccinated naive cattle exposed to B microplus larvae can develop clinical babesiosis before the vaccine has time to produce a protective immunity, which normally occurs within 21 days of vaccination. One infected tick is sufficient to transmit B bovis, but tick infection rates are usually low and the rate of transmission to cattle is therefore usually slow. Studies in Australia found only 0.04% of field ticks to be infected with B bovis in B taurus cattle and even less in the case of B indicus (Mahoney and Mirre 1971; Mahoney et al 1981). Tick infection rates with B bigemina are higher (0.23% in Mahoney and Mirre 1971 study) and transmission rates in this species are therefore higher. 12

14 Anaplasmosis Anaplasma marginale in Australia is usually transmitted by the vector Boophilus microplus (Rogers and Shiels 1979) although mechanical (Ristic 1968) and in utero (Potgieter and Van Rensburg 1987) transmission may also be significant. Transmission of A marginale by the one-host tick B microplus has been shown in pen trials to be both transstadial and intrastadial as this species is incapable of transovarial transmission (Connell 1974; Leatch 1973). Mason and Norval (1981) demonstrated the transfer of larval and adult male B microplus from tick infested to uninfested cattle under field conditions. The transmission of anaplasmosis by interhost transfer of male B microplus from Anaplasma infected to susceptible cattle may be most important because males can survive for periods exceeding two months (Callow 1984). For ticks to transmit the infection, it is therefore necessary to have A marginale infected cattle in close contact with susceptible cattle Breed resistance to tick fever To assess the natural resistance of breeds to tick fever parasites, Bock et al (1997) inoculated 18 month-old naive Bos indicus, cross-bred and Bos taurus steers sequentially with blood infected with B bovis, B bigemina and then A marginale. Criteria to assess reactions after infection included parasitaemia, fever, depression in packed cell volume (PCV) and the need for specific therapy as determined by preset treatment criteria. The treatment proportions are shown in Table 5. Table 5: Proportion of 18 month-old naive steers of differing breeds requiring treatment when inoculated with blood infected with virulent isolates of B bovis, B bigemina, or A marginale. (Adapted from Bock et al 1997) Treatment proportion Group Breed Trial 1 B bovis Trial 2 B bigemina Trial 3 A marginale 1 Pure B indicus 0 / 10 0 / 10 5 / 10 2 half B indicus 3 / 10 0 / 9 7 / 10 3 quarter B indicus 2 / 10 0 / 10 8 / 10 4 Pure B taurus 8 / 10 B indicus breeds almost invariably experience milder primary reactions than B taurus when infected with either species of Babesia (Lohr KF 1973; Anon 1975; Anon 1981; Callow LL 1984). This phenomenon is thought to be a result of the evolutionary relationship between B indicus cattle, the cattle tick and babesia (Dalgliesh RJ 1993). The results in Table 5 confirm information on breed susceptibility to B bovis available in the literature (Callow LL 1984; Daly and Hall 1955; Johnston and Sinclair 1980). The B bovis isolate used induced highly pathogenic infections in B taurus cattle, but the B indicus cattle, in contrast, overcame the infection without difficulty. An important finding is the relative susceptibility of the two crossbred cattle groups with 25% requiring treatment. This suggests that in crossbred cattle the resistance conferred by the B indicus component may not be adequate to prevent economically significant losses due to B bovis infection. Bock et al (1997) found that A marginale infection caused significant disease in all breeds studied (Table 5). This confirmed the work of Wilson et al (1979) and Otim et al (1980) who reported no difference in susceptibility of B taurus and B indicus cross-bred cattle (½ to ¾ B indicus) to A marginale. These reports suggest that A marginale could be a significant cause of disease in B indicus and B indicus crossbred herds. 0 / / 10 13

15 There is some ambiguity in the literature concerning the relative susceptibility of B indicus and B taurus breeds to B bigemina. This parasite is usually of low pathogenicity in Australia and infections are usually not lethal even when fully susceptible cattle are introduced to an enzootic area (Callow LL 1979; Meehan 1969). Daly and Hall (1955) reported no observable difference between the susceptibility of B indicus (Zebu), crossbred (Santa Gertrudis and Africander) and B taurus (British) breed cattle to this parasite during vaccination. Johnston (1967) reported similar findings based on relative parasitaemias in thick blood films from naturally infected B taurus (Hereford) and crossbred (Droughtmaster) cattle. Although Parker et al (1985) demonstrated that pure B indicus were significantly more resistant than B taurus (Shorthorn) cattle, he conceded that the former may not have a distinct advantage in field situations due to the severity of reactions seen in both breeds. Lohr (1973) concluded that B indicus (Sahiwal) cattle were more resistant than B taurus (Charolais). The B bigemina isolate used by Bock et al (1997) produced a mild response in all groups, but pure B indicus cattle as well as crossbreds with half and quarter B indicus were more resistant to challenge than B taurus cattle (Table 5). In a more recent study Bock et al (1999) using a more virulent B bigemina isolate again found pure B indicus cattle and crossbreds to be more resistant to challenge than B taurus cattle (Table 6). The results were generally uniform within groups, but one of the crossbreds was significantly more affected and one of the B taurus animals significantly less affected than group cohorts. This suggests that although B indicus and B indicus cross animals are generally more resistant than B taurus cattle, the susceptibility of individuals within these breeds does vary. This individual variation coupled with the variation of virulence of B bigemina field isolates, may help to explain the ambiguity in the literature with regard to breed susceptibility to B bigemina as well as the occasional severe outbreaks seen in B indicus and B indicus cross cattle. Table 6: Comparison of the innate immunity of steers of differing breed when infected with the virulent Euthella isolate of B bigemina showing pure B indicus and B indicus cross B taurus steers to be more resistant than B taurus steers. (From Bock et al 1999) Breed Mean maximum percent PCV depression Mean total parasitaemia score Maximum temperature rise ( o C) Pure B indicus 22.4 A 6.3 A 0.5 A half B indicus 31.4 B 14.3 A 0.7 A Pure B taurus 50.9 C 32.1 B 1.8 B Least significant difference at % level Values in columns are the group means. Within column means with different superscripts are significantly different to the 5% level of probability. Group 2 had only 6 animals. Johnston et al (1978) showed that immunity of Bos indicus cross cattle to B bovis lasted at least 3 years despite the fact that most of the trial cattle eliminated the infection during that time. Unfortunately, Johnston and coworkers did not monitor antibody levels in the cattle. In a different study, however, Callow et al (1974) showed that indirect fluorescent antibody tests (IFAT) on sera from vaccinated cattle sterilised with Imizol were generally negative within six months of treatment, but this decrease in IFAT titre was not associated with a loss of immunity. In a long-term study of immunity in Bos taurus cattle to B bovis, Mahoney et al (1979a) found that 30% to 50% of the vaccinated or infected trial cattle became seronegative in an indirect haemaglutination test during the trial period. However these cattle were still immune 4 years after initial vaccination or exposure. 14

16 There is also little field evidence of immunity waning to A marginale in Australia, despite the observation that antibody titres will fall off rapidly (author s observations). Animals inoculated with A centrale remain infected for a long time, probably for life (Krigel et al 1992) and there is no indication that the immunity following A centrale infection will wane with time. However, there is some evidence that cattle vaccinated at a young age when non-specific resistance is high, may experience such mild reactions that the immune response is insufficient (Anon 1984). Wilson et al (1980) found that there was no field evidence to suggest that vaccination of young animals is not effective in Australia. Where the A marginale challenge is strong, cross-immunity between the species may not be sufficient. In Australia the A marginale strains appear to be of moderate virulence and the protection provided by A centrale is usually adequate (Anon 1984). So it appears that a persistent, detectable antibody titre is not a prerequisite for immunity. However, it is a very effective indicator of recent infection either naturally or by vaccination Endemic stability Endemic stability is defined as the state where the relationship between host, agent, vector and environment is such that clinical disease occurs rarely or not at all (Perry 1996). Young cattle have higher innate resistance to most tick-borne diseases (Mahoney and Ross 1972) and consequently disease incidence and corresponding mortality are typically lower for this stock class. Because of the low and variable rate of infection in B microplus, tick fever infection does not always follow exposure to ticks. When the number of ticks infected is low, thousands need to bite an animal to ensure it becomes infected. In tick-infested areas, droughts or dipping practices and use of resistant cattle may delay transmission of tick fever parasites for months or even years Age resistance The age of cattle at the time of first exposure to tick fever usually determines the disease outcome. Older cattle will be more severely affected. Calves from immune mothers receive colostral protection against tick fever lasting about 2 months (Callow and Dalgliesh 1982). In most calves an age resistance that lasts a further 4-7 months (Callow 1984) follows this. Resistance then gradually wanes with time and the cattle can become very susceptible to tick fever (Trueman and Blight 1978; Paul et al 1980). Calves exposed to tick fever during the first 6 to 9 months when the age resistance is at its peak, rarely show clinical symptoms and usually develop a solid long-lasting immunity. Cattle not exposed as calves to tick fever may develop severe, life threatening disease if they are exposed later in life, depending on their breed. Mahoney estimated that if at least 75% of calves were exposed to tick fever by 9 months of age the disease incidence would be very low and a state of endemic stability would exist. B bigemina infects more cattle as calves because higher proportions of ticks carry it compared to B bovis (Mahoney and Mirre 1971). Calf exposure is hard to predict for A marginale because tick infection rates are variable and depends on access to infected cattle in the herd, but can be up to 100% (Dallwitz 1987). 15

17 7.4 Models of endemic stability and disease prediction Mahoney model Simple mathematical models can be used to predict the level of endemic stability for tick fever in a herd. Mahoney and Ross (1972) developed a babesiosis model, which utilises the rate in infection in calves (serological). The probability of occurrence of babesiosis outbreaks is directly related to the proportion of non-immune cattle in a herd and to the inoculation rate (Mahoney and Ross 1972; Mahoney 1974). The inoculation rate is the average daily probability of infection and depends on the number of ticks that bite each animal daily, the proportion of ticks infected with the babesia organisms and the proportion of bites from infected ticks that transmit infection. Inoculation rate is related to host infection, which can be measured by serological tests (Mahoney and Ross 1972; Mahoney 1973; Mahoney 1974). If the inoculation rate is high, young animals (less than 9 months old) are exposed to infection before passive protection from maternal antibodies or non-specific immunity wanes and may not experience overt clinical disease when they are exposed to infection as adults. For endemic stability to be achieved Mahoney and Ross (1972) estimated the inoculation rates in B taurus cattle should be between and This is an average tick infestation of greater than 20 adult ticks/day/animal or averages of 1 infective tick bite per calf every 100 to 200 days. For B indicus cattle and their crosses Mahoney (1979) estimated infestations of 40 ticks per day per animal under favourable conditions for tick survival are necessary to achieve endemic stability. The bovine babesiasis inoculation rate model described by Mahoney and Ross (1972), has been used in Australia to predict the proportion of cattle that will become exposed to B bovis after a particular number of years ranging from one to seven years (Mahoney 1973). The model, however, has some limitations in that it is less accurate in B indicus cattle due to the low tick numbers infecting these cattle and their natural resistance Ramsay s disease prediction/vaccination and discounted cashflow analysis model Ramsay (1997) designed an alternative model, which can be used to predict babesia infections in both Bos indicus and Bos taurus cattle. The Ramsay disease prediction/vaccination spreadsheet model involves the use of a method to transform seroprevalence data to incidence risk, which is incorporated into a Markov chain disease prediction model. The Markov model simulates the movement between disease states over time using conditional probability. The disease model is in turn linked to a herd model. The disease model predicts the proportion of animals in each age and sex class that would be affected by different severities of disease. The herd model then converts these proportions to estimates of the number of animals in each disease severity class. The model is not designed to examine epidemic spread of disease between properties and is restricted to endemic disease that cycles within a herd. Predictions are made for eight age classes from zero years to seven years for male and female classes. Because of the lack of a suitable method to classify the severity of disease following infection a simplified method was adopted that divided disease into categories which could be reclassified, as additional information became available. The model includes disease states, which are thought to have an effect on health and production. 16

18 The five states included in the model for infected cattle are: Category 1 Category 2 Category 3 Category 4 Category 5 subclinical disease; mild disease with recovery; severe disease with recovery; acute disease with death; chronic disease with death. The relationships between these states are illustrated in the transition diagram in Figure 1 Vaccination Subclinical, recover Mild, recover Immune Susceptible Infected Severe, recover Acute, die Chronic, die Figure 1: Transition diagram for different states in the disease prediction model including vaccination (Ramsay PhD Thesis 1997) Inputs and outputs of the model and the principles involved in the calculations are illustrated in Figure 2. State probabilities for the various levels of resistance were estimated from trial data and expert opinion (Appendix 3) Disease prediction model Ramsay uses a formula from Lilienfeld and Lilienfeld (1980) to calculate incidence risk from age specific seroprevalence. Incidence risk of seroconversion = 1 (10 (log (1-Pn)/n)) This formula calculates the incidence of seroconversion and where significant mortalities occur with differences between age groups, adjustments need to be made. In the case of tick fever, considerable variation occurs between types of cattle. The disease response is therefore estimated in the Ramsay model for 3 levels of susceptibility to the disease. These susceptibility classes are susceptible (B taurus), intermediate (B taurus cross B indicus) and resistant (B indicus) breeds. Additionally, age has an important effect on susceptibility to disease with animals less than one year old less likely to show clinical signs or die from infection. Disease susceptibility is assessed for two age classes, less than one year and greater than one year old for each susceptibility class. 17

19 Incidence risk of infection = Incidence _ risk _ of _ seroconversion ( p1 + p2 + p3) Where p1 equals the probability of infected animals being in category 1 and p2 equals the probability of infected animals being in category 2 etc. The probability of category 4 and 5 (p4, p5) is not used in the formula as animals that die cannot obviously be subsequently sampled or counted. This gives the probability of progression from susceptible to infected in one year and is calculated from the serological data. The seroprevalence can be taken from survey data for the shire in which a property occurs or more accurately using serology on weaner age cattle from the property being assessed. The proportion of animals in any age group that have become infected in the past year can be calculated as: Proportion infected = proportion susceptible x IR of infection The proportion susceptibles being (1 proportion seropositive) and the proportion of the age group that are in a specific disease category eg Category 4 can be calculated as: Proportion infected for age group = proportion infected x proportion in category 4 The conditional probability of progression from infected to each of the five severities of disease has been estimated from trial data and expert opinion and using this information Ramsay s model simulates the disease occurrence over several years (Appendix 3). The input parameters and flow of the model is illustrated in figure 2. Apparent prevalence Age Test sensitivity and specificity Incidence of infection Proportions of immune animals in different age groups Proportions of susceptible animals (1 - proportion immune) Proportion of new infections at each disease severity for each age group from 0-7 years for the next 8 years Vaccine coverage and efficacy Vaccination strategy Proportions of animals in different disease states for the following breed classes B. indicus - Resistant B. taurus - Susceptible Crosses - Intermediate Number cases of disease avoided at each level of disease severity Production losses avoided due to vaccination Figure 2: Inputs (shown in italics) and outputs of the disease prediction model (adapted from Ramsay, 1997) 18

20 7.5 Assumptions of the disease prediction model The disease prediction model is based on the following assumptions: Immunity following natural infection does not wane It is assumed the carrier state does not affect production Recovery from disease is complete and there is no recurrence of disease in recovered animals An animal that is seropositive for a disease is immune to that disease and an animal that is not seropositive is susceptible The incidence risk of infection is constant between age groups and years and is independent of the sex of the animal Vaccination immunity to tick fever does not affect transmission rate There is ample evidence in the literature that babesia antibodies directed against babesia parasites might not be detected in animals which were originally infected. This is largely due to a time-related decrease in antibody levels in infected cattle. The period of time varies from less than 1 year in Bos indicus cattle to at least 4 years in Bos taurus (Johnston et al 1978; Callow et al 1984; Mahoney et al 1979). It has also been shown that disease transmission and survival of ticks on pasture is dependent on climatic conditions like temperature, humidity and rainfall, which vary between years (Dalgliesh et al 1979; Hartley 1966; de Vos 1979). Incidence risk of infection of babesia is therefore not constant between years. No attempt has been made to estimate what these changes may be, but instead the model is constructed so that incidence risk (that is transition probability to go from susceptible to infected status) can be varied between years and age groups. To enable variation in the incidence risk of infection the model, from the serological data, determines the steady state and then the incidence rate can be manually varied. Age specific data for herd structure, mortality rate and culling rate are put into a herd model, which then maintains a constant herd size and structure. The herd can be actual data or a model herd for the particular region or shire. All cows are culled at 8 years of age and the proportion of males sold for each age group is set. The results of the project surveys and trials are used throughout the costing analysis, and supplemented with expert opinion where data deficiencies exist (Appendix 2) Vaccination strategies The model developed by Ramsay provides a way to determine the effect of vaccination on herd immunity and predict the number of cases of disease that would be prevented by a vaccination program. This allows producers to estimate the benefit of vaccination for particular incidence risks of infection and herd resistance. Two vaccination strategies are used in the model: 19

21 Vaccination program 1 - In this program only animals one year old are vaccinated in each year of the program. Vaccination program 2 - In this program all animals are vaccinated in the first year of the program and animals one year old are vaccinated in subsequent years. The model also allows for the estimation of production loss avoided by vaccination. The production loss is divided into loss due to deaths, weight lost and not regained by the time of sale and reproductive loss. Data for the weight and reproductive loss comes from expert opinion, but mortality is by far the most important production loss parameter for the tick fevers Vaccine coverage and efficacy In the model vaccination coverage which is dependent on mustering rate etc and infectivity of the vaccine as well as vaccine efficacy can be adjusted. In the worked examples we have used 100% musters and the proportion vaccinated reflects infectivity. For the babesia vaccines the infectivity and efficacy are usually set at 99% and 95%, respectively. For the A centrale infectivity is set at 95% and efficacy at 90%. This gives a transition probability from susceptible to immune of 94% and 85.5% for the Babesia and Anaplasma vaccines, respectively Discounted cash flow analysis The aim of discounted cashflow analysis is to reduce the costs and benefits of the control program to a common unit, money at a set point in time, usually the present, by the use of discounting. The Ramsay model generates three commonly used criteria: benefit to cost ratio, net present value and the internal rate of return to compare the performance of the vaccination strategies. The cost of vaccination is estimated in the model using the number of doses used and vaccine prices as well as mustering and application costs. Also the sale prices of cattle and disease and treatment costs are factored into the overall model as outlined in appendix 3. In the worked example, sale prices were taken from the Country Life sales report February Vaccine prices are calculated as the bivalent price of $1.74 per dose with half of this being allocated to B bovis and half to Anaplasma vaccine. It is assumed that vaccination is carried out in conjunction with a routine muster so no additional mustering cost is added. A vaccine application cost of $0.25 per dose was allocated to B bovis and Anaplasma vaccine to cover freight and the labour costs of the actual inoculation. If B bigemina vaccine is used the cost is calculated as 60 cents per dose (rounded up) as it is only available with the other two parasites in the Trivalent vaccine. No application cost for B bigemina vaccine is added in the worked example as this is factored into the other B bovis and A marginale components of the vaccine. The Ramsay model uses as the default discount rates of 4%, 8% or 12%. These rates can be modified and the current discount rate is estimated to be about 7% (Trevor Wilson QDPI economist, personnel communication) The effect of vaccination programs on the economic performance criteria can therefore be rapidly assessed for various levels of age specific seroprevalence (incidence risk of infection) and susceptibility of cattle within different herd profiles. 20

22 7.5.4 Project objectives and findings Serological survey (Objective 1) Objective 1: Perform survey of the current prevalence of the three tick fever parasites (A marginale, B bovis and B bigemina) in cattle on a cross section of properties in northern cattle industry over a period of 3 years. During 1996, we conducted a serological survey to determine the current prevalence of tick fever parasites in weaner cattle in north-west Queensland. The sampling unit was the herd. The herds were randomly selected from a sampling frame of the 530 herds with greater than 1000 head in Burke, Carpenteria, Croydon, Etheridge and tick infested areas of Cloncurry, Dalrymple, McKinley Mt Isa, Richmond and Flinders shires. As herd sizes tend to vary along shire boundaries, herds were stratified by shire and selected accordingly. Cattle numbers in the shire and the number of properties in the shire weighted the number of herds selected per shire. A total of approximately 60 non-vaccinated cattle aged between 6 and 12 months were sampled from each of 116 herds. This age group was selected, as antibody levels to tick fever will not persist in cattle particularly B indicus types. A total of 7067 samples were tested for antibody to B bovis with an ELISA (Molloy et al 1998a), B bigemina ELISA (Molloy et al 1998b) and Anaplasma with a Card Agglutination Test (CAT) (Amerault and Roby 1968). A summary of results are presented in Table 7 Table 7: Summary of 1996 serological survey of herds in NW Queensland Percent of animals seropositive per herd B bovis A marginale B bigemina Shire Herds average range average range average range Burke Croydon Carpenteria Cloncurry Dalrymple Etheridge Flinders Mt Isa McKinley Richmond Total The survey indicates that transmission of both B bovis and A marginale was very low for all shires except Burke and Carpenteria. B bigemina transmission was also generally low, but the range was much broader indicating higher property to property variation. The results are similar to those of smaller pilot surveys and structured surveillance carried out in this area between 1990 and 1995 and some follow-up sampling in 1997 (Table 8). 21